3) assumes statistical independence
among the local test results, the FDR procedure is (as will be illustrated in the following section) approximately valid (while being somewhat conservative) even when those results are strongly correlated, unlike the use of Eq.
So any estimation of probabilities associated with the TC must account for this lack of statistical independence
as well as for the nature of the distributions.
As can be seen in Tables 3 and 4, the assumption of statistical independence is contradicted by the data observed in relation to multiple variables.
Where the assumption of statistical independence is not contradicted, no differences between the distributions exist, namely with none or limited (non-significant) statistical differences, indicating no or limited relationships between the variables.
To further ensure statistical independence
, we randomize the hold-off time between measurements.
The following sections of this paper examine the nature of statistical independence
and test the statistical independence
of the two key variables, the participation rate and the employment rate.
Consequently, the kriged variance for a set of measured (independent) data and some arbitrary but not over-smoothed set of kriged (dependent) data violate the requirement of statistical independence
This condition leads us to a very useful measure for investigating the statistical independence
of the k-tuple consisting of successive terms of a GFSR sequence.
The reason for this is that the latter increases the possibility of the statistical independence
being violated and systematic error being created (McMillan, 1999; Onwuegbuzie, in press).